Single experiment example
Leiv Rønneberg
05/09/2022
In the R package, we’ve attached two example datasets from a large
drug combination screening experiment on diffuse large B-cell lymphoma.
We’ll use these to show some simple use cases of the main functions and
how to interpret the results.
Let’s load in the first example and have a look at it
library(bayesynergy)
data("mathews_DLBCL")
y = mathews_DLBCL[[1]][[1]]
x = mathews_DLBCL[[1]][[2]]
head(cbind(y,x))
## Viability ibrutinib ispinesib
## [1,] 1.2295618 0.0000 0
## [2,] 1.0376006 0.1954 0
## [3,] 1.1813851 0.7812 0
## [4,] 0.5882688 3.1250 0
## [5,] 0.4666700 12.5000 0
## [6,] 0.2869514 50.0000 0
We see that the the measured viability scores are stored in the
vector y, while x is a matrix with two columns
giving the corresponding concentrations where the viability scores were
read off.
Fitting the regression model is simple enough, and can be done on
default settings simply by running the following code (where we add the
names of the drugs involved, the concentration units for plotting
purposes, and calculate the bayes factor).
fit = bayesynergy(y,x, drug_names = c("ibrutinib", "ispinesib"),
units = c("nM","nM"),bayes_factor = T)
##
## SAMPLING FOR MODEL 'gp_grid' NOW (CHAIN 1).
## Chain 1:
## Chain 1: Gradient evaluation took 0.000587 seconds
## Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 5.87 seconds.
## Chain 1: Adjust your expectations accordingly!
## Chain 1:
## Chain 1:
## Chain 1: Iteration: 1 / 2000 [ 0%] (Warmup)
## Chain 1: Iteration: 200 / 2000 [ 10%] (Warmup)
## Chain 1: Iteration: 400 / 2000 [ 20%] (Warmup)
## Chain 1: Iteration: 600 / 2000 [ 30%] (Warmup)
## Chain 1: Iteration: 800 / 2000 [ 40%] (Warmup)
## Chain 1: Iteration: 1000 / 2000 [ 50%] (Warmup)
## Chain 1: Iteration: 1001 / 2000 [ 50%] (Sampling)
## Chain 1: Iteration: 1200 / 2000 [ 60%] (Sampling)
## Chain 1: Iteration: 1400 / 2000 [ 70%] (Sampling)
## Chain 1: Iteration: 1600 / 2000 [ 80%] (Sampling)
## Chain 1: Iteration: 1800 / 2000 [ 90%] (Sampling)
## Chain 1: Iteration: 2000 / 2000 [100%] (Sampling)
## Chain 1:
## Chain 1: Elapsed Time: 18.0561 seconds (Warm-up)
## Chain 1: 17.068 seconds (Sampling)
## Chain 1: 35.1241 seconds (Total)
## Chain 1:
##
## SAMPLING FOR MODEL 'gp_grid' NOW (CHAIN 2).
## Chain 2:
## Chain 2: Gradient evaluation took 0.000316 seconds
## Chain 2: 1000 transitions using 10 leapfrog steps per transition would take 3.16 seconds.
## Chain 2: Adjust your expectations accordingly!
## Chain 2:
## Chain 2:
## Chain 2: Iteration: 1 / 2000 [ 0%] (Warmup)
## Chain 2: Iteration: 200 / 2000 [ 10%] (Warmup)
## Chain 2: Iteration: 400 / 2000 [ 20%] (Warmup)
## Chain 2: Iteration: 600 / 2000 [ 30%] (Warmup)
## Chain 2: Iteration: 800 / 2000 [ 40%] (Warmup)
## Chain 2: Iteration: 1000 / 2000 [ 50%] (Warmup)
## Chain 2: Iteration: 1001 / 2000 [ 50%] (Sampling)
## Chain 2: Iteration: 1200 / 2000 [ 60%] (Sampling)
## Chain 2: Iteration: 1400 / 2000 [ 70%] (Sampling)
## Chain 2: Iteration: 1600 / 2000 [ 80%] (Sampling)
## Chain 2: Iteration: 1800 / 2000 [ 90%] (Sampling)
## Chain 2: Iteration: 2000 / 2000 [100%] (Sampling)
## Chain 2:
## Chain 2: Elapsed Time: 18.6252 seconds (Warm-up)
## Chain 2: 19.1201 seconds (Sampling)
## Chain 2: 37.7454 seconds (Total)
## Chain 2:
##
## SAMPLING FOR MODEL 'gp_grid' NOW (CHAIN 3).
## Chain 3:
## Chain 3: Gradient evaluation took 0.00032 seconds
## Chain 3: 1000 transitions using 10 leapfrog steps per transition would take 3.2 seconds.
## Chain 3: Adjust your expectations accordingly!
## Chain 3:
## Chain 3:
## Chain 3: Iteration: 1 / 2000 [ 0%] (Warmup)
## Chain 3: Iteration: 200 / 2000 [ 10%] (Warmup)
## Chain 3: Iteration: 400 / 2000 [ 20%] (Warmup)
## Chain 3: Iteration: 600 / 2000 [ 30%] (Warmup)
## Chain 3: Iteration: 800 / 2000 [ 40%] (Warmup)
## Chain 3: Iteration: 1000 / 2000 [ 50%] (Warmup)
## Chain 3: Iteration: 1001 / 2000 [ 50%] (Sampling)
## Chain 3: Iteration: 1200 / 2000 [ 60%] (Sampling)
## Chain 3: Iteration: 1400 / 2000 [ 70%] (Sampling)
## Chain 3: Iteration: 1600 / 2000 [ 80%] (Sampling)
## Chain 3: Iteration: 1800 / 2000 [ 90%] (Sampling)
## Chain 3: Iteration: 2000 / 2000 [100%] (Sampling)
## Chain 3:
## Chain 3: Elapsed Time: 18.3892 seconds (Warm-up)
## Chain 3: 18.8873 seconds (Sampling)
## Chain 3: 37.2765 seconds (Total)
## Chain 3:
##
## SAMPLING FOR MODEL 'gp_grid' NOW (CHAIN 4).
## Chain 4:
## Chain 4: Gradient evaluation took 0.000346 seconds
## Chain 4: 1000 transitions using 10 leapfrog steps per transition would take 3.46 seconds.
## Chain 4: Adjust your expectations accordingly!
## Chain 4:
## Chain 4:
## Chain 4: Iteration: 1 / 2000 [ 0%] (Warmup)
## Chain 4: Iteration: 200 / 2000 [ 10%] (Warmup)
## Chain 4: Iteration: 400 / 2000 [ 20%] (Warmup)
## Chain 4: Iteration: 600 / 2000 [ 30%] (Warmup)
## Chain 4: Iteration: 800 / 2000 [ 40%] (Warmup)
## Chain 4: Iteration: 1000 / 2000 [ 50%] (Warmup)
## Chain 4: Iteration: 1001 / 2000 [ 50%] (Sampling)
## Chain 4: Iteration: 1200 / 2000 [ 60%] (Sampling)
## Chain 4: Iteration: 1400 / 2000 [ 70%] (Sampling)
## Chain 4: Iteration: 1600 / 2000 [ 80%] (Sampling)
## Chain 4: Iteration: 1800 / 2000 [ 90%] (Sampling)
## Chain 4: Iteration: 2000 / 2000 [100%] (Sampling)
## Chain 4:
## Chain 4: Elapsed Time: 21.6373 seconds (Warm-up)
## Chain 4: 16.7081 seconds (Sampling)
## Chain 4: 38.3453 seconds (Total)
## Chain 4:
##
## SAMPLING FOR MODEL 'nointeraction' NOW (CHAIN 1).
## Chain 1:
## Chain 1: Gradient evaluation took 0.000126 seconds
## Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 1.26 seconds.
## Chain 1: Adjust your expectations accordingly!
## Chain 1:
## Chain 1:
## Chain 1: Iteration: 1 / 2000 [ 0%] (Warmup)
## Chain 1: Iteration: 200 / 2000 [ 10%] (Warmup)
## Chain 1: Iteration: 400 / 2000 [ 20%] (Warmup)
## Chain 1: Iteration: 600 / 2000 [ 30%] (Warmup)
## Chain 1: Iteration: 800 / 2000 [ 40%] (Warmup)
## Chain 1: Iteration: 1000 / 2000 [ 50%] (Warmup)
## Chain 1: Iteration: 1001 / 2000 [ 50%] (Sampling)
## Chain 1: Iteration: 1200 / 2000 [ 60%] (Sampling)
## Chain 1: Iteration: 1400 / 2000 [ 70%] (Sampling)
## Chain 1: Iteration: 1600 / 2000 [ 80%] (Sampling)
## Chain 1: Iteration: 1800 / 2000 [ 90%] (Sampling)
## Chain 1: Iteration: 2000 / 2000 [100%] (Sampling)
## Chain 1:
## Chain 1: Elapsed Time: 2.75757 seconds (Warm-up)
## Chain 1: 3.04527 seconds (Sampling)
## Chain 1: 5.80284 seconds (Total)
## Chain 1:
##
## SAMPLING FOR MODEL 'nointeraction' NOW (CHAIN 2).
## Chain 2:
## Chain 2: Gradient evaluation took 0.000101 seconds
## Chain 2: 1000 transitions using 10 leapfrog steps per transition would take 1.01 seconds.
## Chain 2: Adjust your expectations accordingly!
## Chain 2:
## Chain 2:
## Chain 2: Iteration: 1 / 2000 [ 0%] (Warmup)
## Chain 2: Iteration: 200 / 2000 [ 10%] (Warmup)
## Chain 2: Iteration: 400 / 2000 [ 20%] (Warmup)
## Chain 2: Iteration: 600 / 2000 [ 30%] (Warmup)
## Chain 2: Iteration: 800 / 2000 [ 40%] (Warmup)
## Chain 2: Iteration: 1000 / 2000 [ 50%] (Warmup)
## Chain 2: Iteration: 1001 / 2000 [ 50%] (Sampling)
## Chain 2: Iteration: 1200 / 2000 [ 60%] (Sampling)
## Chain 2: Iteration: 1400 / 2000 [ 70%] (Sampling)
## Chain 2: Iteration: 1600 / 2000 [ 80%] (Sampling)
## Chain 2: Iteration: 1800 / 2000 [ 90%] (Sampling)
## Chain 2: Iteration: 2000 / 2000 [100%] (Sampling)
## Chain 2:
## Chain 2: Elapsed Time: 2.77665 seconds (Warm-up)
## Chain 2: 2.65496 seconds (Sampling)
## Chain 2: 5.4316 seconds (Total)
## Chain 2:
##
## SAMPLING FOR MODEL 'nointeraction' NOW (CHAIN 3).
## Chain 3:
## Chain 3: Gradient evaluation took 9.6e-05 seconds
## Chain 3: 1000 transitions using 10 leapfrog steps per transition would take 0.96 seconds.
## Chain 3: Adjust your expectations accordingly!
## Chain 3:
## Chain 3:
## Chain 3: Iteration: 1 / 2000 [ 0%] (Warmup)
## Chain 3: Iteration: 200 / 2000 [ 10%] (Warmup)
## Chain 3: Iteration: 400 / 2000 [ 20%] (Warmup)
## Chain 3: Iteration: 600 / 2000 [ 30%] (Warmup)
## Chain 3: Iteration: 800 / 2000 [ 40%] (Warmup)
## Chain 3: Iteration: 1000 / 2000 [ 50%] (Warmup)
## Chain 3: Iteration: 1001 / 2000 [ 50%] (Sampling)
## Chain 3: Iteration: 1200 / 2000 [ 60%] (Sampling)
## Chain 3: Iteration: 1400 / 2000 [ 70%] (Sampling)
## Chain 3: Iteration: 1600 / 2000 [ 80%] (Sampling)
## Chain 3: Iteration: 1800 / 2000 [ 90%] (Sampling)
## Chain 3: Iteration: 2000 / 2000 [100%] (Sampling)
## Chain 3:
## Chain 3: Elapsed Time: 2.42067 seconds (Warm-up)
## Chain 3: 2.25905 seconds (Sampling)
## Chain 3: 4.67972 seconds (Total)
## Chain 3:
##
## SAMPLING FOR MODEL 'nointeraction' NOW (CHAIN 4).
## Chain 4:
## Chain 4: Gradient evaluation took 9.7e-05 seconds
## Chain 4: 1000 transitions using 10 leapfrog steps per transition would take 0.97 seconds.
## Chain 4: Adjust your expectations accordingly!
## Chain 4:
## Chain 4:
## Chain 4: Iteration: 1 / 2000 [ 0%] (Warmup)
## Chain 4: Iteration: 200 / 2000 [ 10%] (Warmup)
## Chain 4: Iteration: 400 / 2000 [ 20%] (Warmup)
## Chain 4: Iteration: 600 / 2000 [ 30%] (Warmup)
## Chain 4: Iteration: 800 / 2000 [ 40%] (Warmup)
## Chain 4: Iteration: 1000 / 2000 [ 50%] (Warmup)
## Chain 4: Iteration: 1001 / 2000 [ 50%] (Sampling)
## Chain 4: Iteration: 1200 / 2000 [ 60%] (Sampling)
## Chain 4: Iteration: 1400 / 2000 [ 70%] (Sampling)
## Chain 4: Iteration: 1600 / 2000 [ 80%] (Sampling)
## Chain 4: Iteration: 1800 / 2000 [ 90%] (Sampling)
## Chain 4: Iteration: 2000 / 2000 [100%] (Sampling)
## Chain 4:
## Chain 4: Elapsed Time: 2.62844 seconds (Warm-up)
## Chain 4: 2.63523 seconds (Sampling)
## Chain 4: 5.26368 seconds (Total)
## Chain 4:
## Calculating the Bayes factor
The resulting model can be summarised by running
## mean se_mean sd 2.5% 50% 97.5% n_eff Rhat
## la_1[1] 0.3364 0.001451 0.0698 1.70e-01 3.46e-01 0.451 2315 1.001
## la_2[1] 0.3747 0.009895 0.0901 3.53e-02 3.96e-01 0.456 83 1.036
## log10_ec50_1 0.4739 0.003236 0.1498 2.46e-01 4.45e-01 0.837 2142 1.002
## log10_ec50_2 -1.0422 0.018874 0.9771 -3.21e+00 -9.20e-01 0.485 2680 1.001
## slope_1 2.0326 0.016696 0.9064 9.12e-01 1.81e+00 4.365 2947 1.000
## slope_2 1.4275 0.039567 1.1476 5.92e-02 1.16e+00 4.286 841 1.006
## ell 3.0534 0.030429 1.5102 1.26e+00 2.72e+00 6.884 2463 1.000
## sigma_f 0.8050 0.014937 0.7859 1.67e-01 5.82e-01 2.778 2769 1.000
## s 0.0971 0.000304 0.0150 7.26e-02 9.52e-02 0.132 2448 0.999
## dss_1 33.5756 0.043330 2.9606 2.76e+01 3.36e+01 39.480 4669 1.000
## dss_2 59.2931 0.051533 2.7395 5.36e+01 5.94e+01 64.490 2826 1.002
## rVUS_f 82.7355 0.013679 0.8862 8.09e+01 8.28e+01 84.436 4197 1.000
## rVUS_p0 72.9829 0.037589 2.1723 6.86e+01 7.30e+01 77.107 3340 1.000
## VUS_Delta -9.7526 0.040124 2.3431 -1.45e+01 -9.67e+00 -5.288 3410 1.000
## VUS_syn -9.7945 0.039432 2.3045 -1.45e+01 -9.69e+00 -5.420 3416 1.000
## VUS_ant 0.0419 0.001836 0.1053 5.07e-06 8.11e-05 0.340 3292 1.000
##
## log-Pseudo Marginal Likelihood (LPML) = 52.0628
## Estimated Bayes factor in favor of full model over non-interaction only model: 35.02581
which gives posterior summaries of the parameters of the model.
In addition, the model calculates summary statistics of the
monotherapy curves and the dose-response surface including drug
sensitivity scores (DSS) for the two drugs in question, as well as the
volumes that capture the notion of efficacy (rVUS_f),
interaction (VUS_Delta), synergy (VUS_syn) and
interaction (VUS_ant).
As indicated, the total combined drug efficacy is around 80%
(rVUS_f), of which around 70 percentage points can be
attributed to \(p_0\)
(rVUS_p0), leaving room for 10 percentage points worth of
synergy (VUS_syn). We can also note that the model is
fairly certain of this effect, with a 95% credible interval given as
(-14.493, -5.42). The certainty of this is also verified by the Bayes
factor, which at 35.03 indicates strong evidence of an interaction
effect present in the model.
Visualization
3D interactive plots
The package can also generate 3D interactive plots by setting
plot3D = T. These are displayed as following using the
plotly library (Plotly Technologies Inc.
(2015)).
References
Plotly Technologies Inc. 2015.
“Collaborative Data
Science.” Montreal, QC: Plotly Technologies Inc. 2015.
https://plot.ly.
